身份验证 (Authentication)

所有请求都需要 dLazy API key。推荐使用 dlazy login 完成登录：

dlazy login

该命令使用设备码流程（远程终端也可用），登录成功后 自动把 API key 写入本地 CLI 配置，无需手动复制粘贴。

备选：手动设置 API Key

如果你已有 API key，也可以直接保存：

dlazy auth set YOUR_API_KEY

CLI 会把 key 保存在你的用户配置目录（macOS/Linux 上为 ~/.dlazy/config.json，Windows 上为 %USERPROFILE%\.dlazy\config.json），文件权限仅限当前操作系统用户访问。你也可以用 DLAZY_API_KEY 环境变量按次传入。

手动获取 API Key

1. 登录或在 dlazy.com 创建账号
2. 访问 dlazy.com/dashboard/organization/api-key
3. 复制 API Key 区域显示的密钥

每个 key 都属于你自己的 dLazy 组织，可在同一控制面板随时轮换或吊销。

关于与来源 (Provenance)

• CLI 源代码: github.com/dlazyai/cli
• 维护者: dlazyai
• npm 包名: @dlazy/cli（本技能 install 字段固定到 1.0.9 版本）
• 官网: dlazy.com

如果你不希望在系统上长期保留一个全局 CLI，可以按需运行：

npx @dlazy/cli@1.0.9 <command>

如选择全局安装，技能的 metadata.clawdbot.install 字段已固定到 npm install -g @dlazy/cli@1.0.9。安装前建议先到 GitHub 仓库审阅源码。

工作原理 (How It Works)

此技能是 dLazy 托管 API 的轻量封装。调用时：

• 你提供的提示词与参数会发送到 dLazy API（api.dlazy.com）进行推理。
• 传入图像 / 视频 / 音频字段的本地文件路径会被 CLI 上传到 dLazy 媒体存储（files.dlazy.com），以便模型读取 —— 与任何云端生成 API 的流程一致。
• API 返回的生成结果 URL 由 files.dlazy.com 托管。

这是标准的 SaaS 调用模式；技能本身不会越权访问网络或文件系统，所有动作都由 dLazy CLI 完成。

Logo & Branding Professional Design System

English · 中文

A professional pipeline for building everything from a core mark to a complete brand visual system, ensuring creative quality, execution consistency, and shippable delivery.

Core Principles

1. Stepwise execution
   • Use a progressive flow; never dump large outputs in one shot.
   • When a single plan exceeds 2 images, submit the execution plan and wait for confirmation before generating.
   • Upstream assets must be confirmed before downstream applications proceed.
2. Object consistency
   • The same brand element stays strictly consistent across all applications.
   • When deriving variants from an existing image, the goal is to replicate consistency, not to reinterpret.
3. User assets first
   • Reference images, logos, and assets provided by the user have the highest priority.
   • Strictly separate visual style from textual content; the brand name uses what the user specifies.

Execution Framework

Step 0: Task Planning (Mandatory)

Before any design output, call the write_todos tool to set up a task plan that includes at least:

• Gather requirements and boundaries
• Design the core logo concept
• Generate derivatives and the brand system
• Phase confirmation and iterative correction

Execution rules:

• Keep only one task in_progress; the rest are pending.
• Update write_todos status as soon as each phase finishes.
• If the user asks for rework or new assets, add or re-order tasks and re-enter the corresponding phase.

Step A: Gather Requirements

You must confirm:

• Brand name
• Industry or product category
• Component scope (logo only, or with derivatives)

Component definitions:

• Logo: the core brand mark
• Derivatives: real-world applications based on the brand mark

Step B: Establish the Core Mark

New mark design follows:

• Single focus: one central element, no background or decorative interference
• Symbolic abstraction: distill to the most recognizable form
• Clear silhouette: high-contrast outline, well-defined edges
• Default 1K resolution

Concept rules:

• For open-ended requests, default to 4 distinct design directions
• Present each concept individually, not as a tile grid
• Each concept comes with a 100–200 word design rationale
• Wait for user confirmation before proceeding

Step C: Create Derivatives

Possible directions:

1. Mark variants
   • Structural variants: lettermark, wordmark, symbol mark, combination mark
   • Layout or orientation variants
   • Color variants
   • Size and use-case variants
2. Brand patterns and supporting graphics
   • Generate independent, reusable graphic assets
3. Physical applications
   • Business cards and stationery
   • Packaging and merchandise
   • Environmental signage and wayfinding
4. Social-media assets
   • Generate ready-to-post final images
5. Brand identity system
   • Brand core (story, pillars, voice)
   • Logo system
   • Color system
   • Typography system
   • Visual language
   • Application examples

Application design principles:

• Place the mark with restraint; avoid oversized, dead-center placements
• Deconstruct brand elements into patterns, icons, color blocks
• Use whitespace to organize visual hierarchy
• Emphasize material and texture experience

Phase Confirmation and Rollback Rules

• Confirm at the end of every step before continuing.
• When the user adjusts an intermediate result, return to the corresponding upstream step first, then sync downstream assets.
• If the user supplies new reference assets, immediately reset the consistency baseline and unify subsequent outputs.

Output Format

• Requirements summary
• Current-phase deliverable
• Next-step plan and items awaiting confirmation
• Current todo status (phase, completed, pending)

Key Value

• Progressive workflow that lowers rework cost
• Multi-direction concept exploration that improves decision quality
• End-to-end consistency control that strengthens brand recognition
• Integrated delivery from logo to brand system

🛠️ 执行与生成指南 (CRITICAL EXECUTION INSTRUCTIONS)

你是可以执行终端命令的智能 Agent！

【严格禁止行为】

• 严禁：将提示词保存到任何文件中（如 txt, md）。
• 严禁：要求用户自己去第三方平台（如 Midjourney）生成图片。
• 严禁：一次性批量生成所有图片，或一次性执行多个命令。

【必须遵循的交互与执行流程】 你必须严格分步执行，并在每一步停下来等待用户回复：

1. 第一步：主动收集需求。当用户提出需求时，不要做任何设计和生成，先向用户提问（如产品特点、目标人群、想要几张图等）。必须等待用户回答。
2. 第二步：输出草案并请求确认。根据用户的回答，制定套图计划，并输出第一张图的提示词草案。询问用户：“是否确认这个提示词，可以开始生成第一张图了吗？” 必须等待用户回答“确认”。
3. 第三步：单次执行终端命令。用户确认后，你必须使用终端执行命令（如 dlazy seedream-4.5 --prompt "..."），每次只能执行一个生成命令。重要：必须使用同步命令，绝不要在命令末尾加 &，绝不要使用 &&，这是在 Windows PowerShell 下运行！
4. 第四步：交付与循环。命令返回结果后，把图片 URL 发给用户，并询问“对这张满意吗？我们可以继续生成下一张了吗？”。收到确认后再继续下一步。